PCN38 Impact of Biomarker Testing on Clinical and Economic Outcomes in Advanced Cancer Treatment

VALUE IN HEALTH 15 (2012) A277–A575

patients included in these studies was 118 patients (min: 30, max: 724). On average, median PFS/TTP was 14.0 months (sd⫽12.4) and median OS was 35.0 months (sd⫽31.2). Results of the correlation analysis indicated that median PFS/TTP is highly correlated with median OS, with a Spearman’s correlation coefficient of 0.813 (pⱕ0.001). A significant correlation between median PFS/TTP and median OS was observed in the second-line and subsequent-line therapies, but not in the first-line setting. CONCLUSIONS: The present results demonstrate a very strong correlation between median PFS/TTP and median OS in the context of CLL, which reinforce the hypothesis that PFS/TTP would be adequate surrogate endpoints for OS in this cancer setting. PCN35 SURVIVAL ESTIMATES OF SIX MAJOR ADVANCED AND METASTATIC SOLID TUMORS IN EUROPE: A CANCER REGISTRY AND LITERATURE REVIEW Moulard O1, Mehta J2, Rose M3, Olivares R3, Hamed A2, Eckert L1 1 Sanofi, Chilly-Mazarin, France, France, 2Sanofi, Cambridge, MA, USA, 3Sanofi, Chilly-Mazarin, France, France

OBJECTIVES: Survival is a key index of the overall effectiveness of health services in the management of patients with cancer. Epidemiological data on cancer survival in unselected patient populations in Europe are sparse and incomplete. Prognosis is directly related to staging of cancer at diagnosis. Our objective is to evaluate variations in survival by using country registry data and published populationbased studies for six tumors: lung, bladder, colorectal, ovarian, pancreatic and gastric. METHODS: Cancer data from nine registries (in France, Italy, Germany, The Netherlands, Austria, Norway, Ireland, Finland, UK) in Europe were used along with literature review using Embase for the period 2000-2012. Relative survival (RS) estimates for various time periods were summarized for stage III and IV for each tumor type. RESULTS: Marked differences in population based survival across Europe were found which could be explained in part by variation in terms of time period and staging definition but also by other factors such as clinical practice, histological type, age, death ascertainment and co-morbidities. 5-year relative survival (RS) in stage III varies greatly according to tumor site with improvement observed over time. 5-year RS rates ranged from 4-24% for lung, 20-45% for bladder, 40-66% for colorectal, 20-61% for ovary, 8% for pancreas and 13-35% for gastric. Survival in metastatic cancer remains low. 1 year RS rates for Stage IV cancers were 7-38% for lung, 24-56% for bladder, 19-81% for colorectal, 34-77% for ovary, 6-16% for pancreas and 6-34% for gastric. CONCLUSIONS: Variations of survival in unselected patient populations are due to differences in patient characteristics, diagnostic codification, and treatment modalities. Future work should validate these findings as registry data is inherent with limitations such as variation in follow up time periods, disease coding and definition of survival. More efforts should focus on standardization of data capturing processes across European cancer registries. CANCER - Cost Studies PCN36 VINORELBINE IN PATIENTS WITH NON-SMALL LUNG CANCER: BUDGET IMPACT ANALYSIS Vitezic D1, Hadzic-Kostrencic C2, Petkovic M1, Dobrila-Dintinjana R1, Vitezica P2, Mrsic Pelcic J3 University of Rijeka Medical School and University Hospital Centre Rijeka, Rijeka, Croatia, 2Pliva Croatia Ltd., Zagreb, Croatia, 3University of Rijeka, Rijeka, Croatia 1

OBJECTIVES: Vinorelbine in non-small lung cancer treatment was not available in Croatia through the Croatian National Health Insurance (CNHI) Positive Drugs List (PDL). The aim of our study was to analyse a possible impact of vinorelbine inclusion to the CNHI PDL according to the budget impact analysis (BIA). METHODS: A clinical guideline for vinorelbine reimbursement by CNHI is proposed according to evidence-based medical criteria and international guidelines. We have calculated the number of patients, potential candidates for vinorelbine treatment, and we have also developed BIA model for three-years period after the drug reimbursement. The share of vinorelbine has been estimated using market data from other countries and the price of vinorelbine has been calculated according to the Croatian MoH Pricing Ordinance. The total costs for CNHI have been calculated using a referent scenario (without vinorelbine) and a scenario with vinorelbine reimbursement. Monte Carlo simulation has been performed too. RESULTS: The total number of patients in non-small cell lung cancer stage III and IV is estimated to be 500-550 per year. Referent treatment costs are between 1,705 EUR and 10,974 EUR per patient (in case of biological treatment usage). Vinorelbine costs are 1,044 EUR (monotherapy) and 1,407 EUR (combination with cisplatin). Monte Carlo simulation showed annual cost of 1,266,794 EUR in scenario without vinorelbine and declining costs in scenario with vinorelbine from 77,851 EUR less in the first year, 158,611 EUR in the second and 237,785 EUR in the third year. Cumulative three-year savings with vinorelbine introduction to the CNHI PDL are 474,321 EUR. CONCLUSIONS: The inclusion of vinorelbine to national reimbursement list for patients with nonsmall lung cancer demonstrates significant savings and for patients provides additional clinical benefits as a new treatment option. PCN37 A PHARMACOECONOMIC ANALYSIS OF THE IMPLEMENTATION OF GEFITINIB IN THE TREATMENT OF METASTASIZED NON-SMALL CELL LUNG CANCER (NSCLC) WITH PROVEN EGFR-TK MUTATIONS Culig J, Leppée M, Skaron-Jakobovic N Andrija Stampar Institute of Public Health, Zagreb, Croatia

OBJECTIVES: In the treatment of locally advanced and metastasized stage of nonsmall cell lung cancer (NSCLC), the first line of medication is comprised of a combination of cisplatin and gemcitabine. The second line of treatment is the combi-

A415

nation of carboplatin and paclitaxel, and the third line is pure erlotinib. A new treatment scenario has been designed which uses the new medication, gefitinib, that acts as an inhibitor of the epidermal growth factor receptor tyrosine-kinase (EGFR-TK), which blocks growth signals and prevents the subsistence of cancer cells. The aim of this study is to assess the potential financial impact of the introduction of gefitinib into the treatment of metastasized NSCLC. METHODS: The study conducted a budget impact analysis, which was meant to determine the difference in cost between the treatment with gefitinib and the established treatment, so as to provide justification for the introduction of the new drug. RESULTS: Direct costs per patient with a proven positive EGFR mutation (5%) treated with the new drug gefitinib, amount to 171.194,00 HRK, while costs for other patients treated with the established medication, amount to 168.793,90 HRK, which shows that the total cost for the new treatment scenario is somewhat higher. The total direct cost for an established treatment scenario is 139.236.532, 10 HRK, while the new scenario amounts to 142.270.005,50 HRK. Scientific observations point to the fact that gefitinib has a higher drug tolerance and better effectiveness among patients diagnosed with an EGFR mutation. CONCLUSIONS: A budget impact analysis shows that 28 patients will be treated with gefitinib in the year 2012, and 96 in the year 2014, which will increase the budget impact by 3%. However, a lowering in administration, patient monitoring and side-effects treatment costs are expected, making the total budget impact of the gefitinib introduction into treatments practically neutral (0,07%). PCN38 IMPACT OF BIOMARKER TESTING ON CLINICAL AND ECONOMIC OUTCOMES IN ADVANCED CANCER TREATMENT Sorensen S1, Pan F1, Knopf KB2 1 United BioSource Corporation, Bethesda, MD, USA, 2Pacific Hematology Oncology Associates, San Francisco, CA, USA

OBJECTIVES: Oncology treatments are often expensive and result in toxicities while not all patients may experience palliative or survival benefits. A growing trend is to utilize biomarkers to better identify treatment responders. In this study, the clinical and economic impact of biomarker testing in advanced non-small cell lung cancer (NSCLC) treatment was explored. METHODS: A simulation model that followed NSCLC patients from start of an investigational treatment, through disease progression and death was developed. The model assessed several biomarker strategies including timing of biomarker assessments and predictive value of the test categorized as true positive (remain on drug), false positive (remain on drug but receive no benefit) and test negative (assume physicians withdrew treatment). Drug related toxicities were defined as immediate or delayed onset to account for differences over time. Disease progression and survival were based on published trials. A hazard ratio of 0.8 for survival and an assumed cost of $60,000 with the investigational agent over standard of care were used. Costs were from US public available sources. Key model outcomes include costs, progression free survival (PFS), overall survival (OS) and quality adjusted survival. RESULTS: The biomarker strategy reduced the cost of treatment while providing PFS and OS similar to the non-biomarker group. A reduction in total costs of 15% was seen, mainly in treatment costs reductions. The biomarker strategy resulted in a 9% improvement in quality adjusted survival due to fewer toxicities associated with avoided nonresponsive treatment. The results vary depending on the accuracy of the test with a trade-off in early testing being less accurate but accruing less treatment costs. CONCLUSIONS: Accurate biomarker tests will not only provide clinical benefit, but significantly reduce the economic burden of payers. It is important to further explore biomarker tests in the economic evaluation of oncology treatments. PCN39 POTENTIAL TIME AND COST SAVINGS WITH HERCEPTIN (TRASTUZUMAB) SUBCUTANEOUS (SC) INJECTION VERSUS HERCEPTIN INTRAVENOUS (IV) INFUSION: RESULTS FROM THREE DIFFERENT ENGLISH PATIENT SETTINGS Samanta K, Moore L, Jones G, Evason J, Owen G Roche Products Limited, Welwyn Garden City, UK

OBJECTIVES: In 2010, there were estimated to be in excess of 75,000 Herceptin IV infusions in England, putting ever-increasing demands on hospital resource. This analysis aims to estimate the change in resource use and subsequent time and cost savings that could result from using Herceptin SC in three different English settings; (1) hospital, (2) community and (3) patient self-administration at home versus Herceptin IV in the hospital. METHODS: Routine data on Herceptin IV-related hospital use and processes was collected using the Chemotherapy Capacity Planning Tool. Interviews were conducted with a number of English centres in order to determine likely changes to key delivery inputs as a result of using Herceptin SC. An economic model was developed to estimate average time and cost savings from delivering Herceptin SC in the three settings assuming 200 patients are given a full treatment course of Herceptin and the capacity of a hospital is 10 IV chairs running 2 full days Herceptin clinics per week. Unit costs were taken from the Personal Social Service Research Unit. RESULTS: A 100% switch of 200 Herceptin patients to SC leads to a fall in total expenditure of £271,000, £1,200,000 and £1,500,000 in settings (1), (2), and (3) respectively. This is largely due to reductions in resource costs from 2,937 hours of IV chair time saved in all settings, 2,691 nurse hours saved in (1) and (2) and 5,712 in (3), and 2,102 pharmacist technician hours saved in (1) and (2) and 4,532 in (3). CONCLUSIONS: A switch from hospital-based IV to SC Herceptin leads to substantial time and cost savings whether delivered in the hospital, in the community or via patient self-administration. Therefore if Herceptin patients received SC administration instead of IV, this could potentially increase hospitals’ throughput and overall efficiency. Any savings could be reinvested elsewhere in order to improve overall patient care.